No, you can't really build it because it's ambiguous.
I entered it into LTSpice, guessing where the grounds and commons should go.
Firstly, consider the other side of each power supply is ground. It can be analysed as I described in my previous post. I added V4 and R3 to prove the potential divider is equivalent to a 500 Ohm resistor to 0V, as far as the 5V PSU is concerned. Note the currents are negative, because LTSpice shows them as the current is being taken from the PSU, i.e. the batteries are being discharged.
ground problem 1a.asc (1.2 kB - downloaded 32 times.)
Secondly, assuming the other side of the 5V rail is ground and the 12V power supplies are floating, with the other sides connected to a different reference, hence the different symbol. No current flows out of the 5V rail, which lifts both 12V supplies up by 5V. This was what I assumed, when I initially replied to the thread.
ground problem 1b.asc (0.97 kB - downloaded 21 times.)
Now let's change the 5V rail to 100V. The 12V rails shift up by 95V, giving 112V and 88V. I stuffed up on my arithmetic on my previous post.
ground problem 1c.asc (0.97 kB - downloaded 28 times.)